Unseen Seer Advice

[JohnDaBarr]

I am currently playing a Rog/Wiz multiclass and will probably prestige into an Unseen Seer, so I am wondering should I take all 10 level of the class? It seems to me that last 2 level don't really offer anything useful.

Also I will probably take 5 lvl's of Wiz to get that extra feat, but I am unsure should I also take another 5 lvl's of Wiz for another feat after I am done with the prestige class? Those 5 lvl's for a single feat at that point of the game don't look like a good investment.
Besides Arcane Trickster and Spellwarp Sniper are there any other roguish arcane classes? (Yes, I am aware of the class Incantrix existence! :smallbiggrin:)

As for Unseen Seer's extra divination spells I decided to take Divine Insight and Hunters Eye, but as for the third one I am somewhat undecided and could use some advice and input.

[Urpriest]

I am currently playing a Rog/Wiz multiclass and will probably prestige into an Unseen Seer, so I am wondering should I take all 10 level of the class? It seems to me that last 2 level don't really offer anything useful.

Also I will probably take 5 lvl's of Wiz to get that extra feat, but I am unsure should I also take another 5 lvl's of Wiz for another feat after I am done with the prestige class? Those 5 lvl's for a single feat at that point of the game don't look like a good investment.
Besides Arcane Trickster and Spellwarp Sniper are there any other roguish arcane classes? (Yes, I am aware of the class Incantrix existence! :smallbiggrin:)

As for Unseen Seer's extra divination spells I decided to take Divine Insight and Hunters Eye, but as for the third one I am somewhat undecided and could use some advice and input.

You get another die of Sneak Attack at level 10, that seems worth it.

Instead of another 5 levels of Wizard, I'd definitely advise Arcane Trickster or Spellwarp Sniper. There are a few other roguish arcane classes (Daggerspell Mage is the one that comes to mind at present) but they all lose caster levels.

For your final divination, something that lets you sneak attack something normally immune (Gravestrike or Vinestrike) might be handy.

[ben-zayb]

I suggest taking the Spontaneous Divination ACF instead of the 5th level bonus feat, especially with Divine Insight in your known spells.

[JohnDaBarr]

You get another die of Sneak Attack at level 10, that seems worth it.

Instead of another 5 levels of Wizard, I'd definitely advise Arcane Trickster or Spellwarp Sniper. There are a few other roguish arcane classes (Daggerspell Mage is the one that comes to mind at present) but they all lose caster levels.

For your final divination, something that lets you sneak attack something normally immune (Gravestrike or Vinestrike) might be handy.

Thx, I'll probably go with the Arcane Trickster and Gravestrike sounds like a solid option.



I suggest taking the Spontaneous Divination ACF instead of the 5th level bonus feat, especially with Divine Insight in your known spells.

Well that certainly sounds interesting, thx on the tip.

[Chronos]

You also get more levels of full casting advancement and good skills. Really, if you're interested in US at all, I'm not sure why you wouldn't take it fully.

[thethird]

Two words: Choose Destiny
Three words: Races of Destiny
Four words: Best Divination Spell Ever
Five words: Seriously you should pick it
Etcetera

[nedz]

Two words: Choose Destiny
Three words: Races of Destiny
Four words: Best Divination Spell Ever
Five words: Seriously you should pick it
Etcetera

You are aware that on average this is only +2.5 ?

That said it does last long enough, and re-rolling 1's is golden.

Also, it's quite high level.

[Fuzzy McCoy]

I'm going to suggest a slightly different option: Divine Oracle. Four levels gets you the Oracle domain (+2 to divination spells with commune on the list), evasion, and uncanny dodge. Plus, highly appropriate!

[nedz]

Divine Oracle is an excellent class and stacks nicely with the Divination boosts of US — but the skills are poor and Mr Rogue probably already has Evasion and Uncanny Dodge. So, a 1 level dip for a domain maybe ? Though I'm not sure it's worth the cost: trading 8 ranks (in likely a cross class skill) and a feat for a domain ?

[A_S]

You are aware that on average this is only +2.5 ?

That said it does last long enough, and re-rolling 1's is golden.

Also, it's quite high level.
This is a misleading characterization of how good 2d20-pick-best really is. The reason is that d20 rolls in D&D are binary (they're either good enough or they're not), rather than linearly scaling (like damage, where rolling twice as much is twice as good). Thus, the appropriate characterization of how good 2d20-pick-best is is not "how much higher is your roll on average," it's "what is the increase in probability that at least one of my rolls is good enough."

This increase depends on the DC of the check compared to your bonus, but it's pretty simple to calculate; the probability of making the check on two dice is just the joint probability of hitting the DC on two independent rolls, which is:

1 - (1 - P)2

...where P is the probability of success on a single roll.

-----



For a check where you have a 50% chance of success on a single roll (you need to roll an 11 or higher), rolling two dice gives you a 75% chance of success, half again what you would otherwise get.

For a check where you have a 95% chance of success on a single roll (you only fail on a natural 1), rolling two dice gives you a 99.75% chance of success (as you pointed out, this is very useful for avoiding unlucky auto-failure of important saves).

For a check where you have a 5% chance of success on a single roll (you need a natural 20), rolling two dice gives you a 9.75% chance of success, nearly twice as good.

I think it's clear from those examples why 2d20-pick-best is a lot better than "only 2.5 higher on average" makes it sound.

[Fuzzy McCoy]

Divine Oracle is an excellent class and stacks nicely with the Divination boosts of US — but the skills are poor and Mr Rogue probably already has Evasion and Uncanny Dodge. So, a 1 level dip for a domain maybe ? Though I'm not sure it's worth the cost: trading 8 ranks (in likely a cross class skill) and a feat for a domain ?

To be fair, I was assuming an optimal build of Rogue 1/Wizard 4-5/USS X, with the able learner feat. If that's not the case then Divine Oracle loses some value. How much depends entirely on how many rogue levels are taken.

[Kraken]

I think it's clear from those examples why 2d20-pick-best is a lot better than "only 2.5 higher on average" makes it sound.

I agree with this sentiment very strongly, I don't have any math to back this up, but my experiences at the table with 'better of 2d20' mechanics in general has gotten me to seek them out when I can reliably get them.

[JohnDaBarr]

Choose Destiny sounds like a useful spell but is a bit to high/late to take, and that would require me to postpone last lvl's to much.

As for the Divine Oracle, that particular class doesn't fit my character mechanically and fluffwise, but thx anyway for the suggestion.

Also is Spellwarp Sniper worth taking the Point Blank feat? My build is somewhat feat starved, as most 3.5 build surly are, so I am not sure should I go that way.

[thethird]

You are aware that on average this is only +2.5 ?

That said it does last long enough, and re-rolling 1's is golden.

Also, it's quite high level.

Objection! A_S said it better though so I'm not going to bother :smallsmile:.


This is a misleading characterization of how good 2d20-pick-best really is. The reason is that d20 rolls in D&D are binary (they're either good enough or they're not), rather than linearly scaling (like damage, where rolling twice as much is twice as good). Thus, the appropriate characterization of how good 2d20-pick-best is is not "how much higher is your roll on average," it's "what is the increase in probability that at least one of my rolls is good enough."

This increase depends on the DC of the check compared to your bonus, but it's pretty simple to calculate; the probability of making the check on two dice is just the joint probability of hitting the DC on two independent rolls, which is:

1 - (1 - P)2

...where P is the probability of success on a single roll.

-----



For a check where you have a 50% chance of success on a single roll (you need to roll an 11 or higher), rolling two dice gives you a 75% chance of success, half again what you would otherwise get.

For a check where you have a 95% chance of success on a single roll (you only fail on a natural 1), rolling two dice gives you a 99.75% chance of success (as you pointed out, this is very useful for avoiding unlucky auto-failure of important saves).

For a check where you have a 5% chance of success on a single roll (you need a natural 20), rolling two dice gives you a 9.75% chance of success, nearly twice as good.

I think it's clear from those examples why 2d20-pick-best is a lot better than "only 2.5 higher on average" makes it sound.

Thanks for the explanation examples A_S that was pretty cool and geeky at the same time. Have an internet cookie.


Choose Destiny sounds like a useful spell but is a bit to high/late to take, and that would require me to postpone last lvl's to much.

As for the Divine Oracle, that particular class doesn't fit my character mechanically and fluffwise, but thx anyway for the suggestion.

Also is Spellwarp Sniper worth taking the Point Blank feat? My build is somewhat feat starved, as most 3.5 build surly are, so I am not sure should I go that way.

Fair enough. I'm just the guy who likes to go full unseen seer for divination purposes more than rogue-y purposes.

My typical unseen seer build looks like:
Human
Spellthief 1 / Wizard 5 (with spontaneous divination) / Incantatrix 3 / Unseen Seer 5 / Prestige Bard 1 / Unseen Seer +5

Feats being: Able learner, Extend Spell, Improved Initiative (Martial Wizard ACF), Knowledge Devotion, Iron Will (Otyugh Hole), Practiced Spellcaster (Wizard), Persist Spell (Incantatrix bonus feat), Versatile Spellcaster, Master Spellthief, Improved Familiar (Imp, Pseudodragon or Lantern Archon), Obtain Familiar

Advanced learning being: Hunter's Eyes, That art thou and Choose Destiny

Persisting Hunter's eyes, Sniper's Shot, That art thou, Choose Destiny, Cloud of Knives and other archery related spells in the character +2 familiars

Edit: For another unusual recomendation That art thou 1 round/level +20 search, spot, listen, seeing/feeling everything that any other person, creature, object sees/feels in 30 ft. And cannot be flanked unless every other creature in 30ft is flanked.

[nedz]

My maths is correct — we did lots of calculations in a thread on this very site a couple of years ago; albeit we were looking at the feat which allows you to re-roll initiative and comparing that to Improved initiative.

I am aware that it squares the error probability but that said: At the level you get to cast 9th level spells it's only likely to be relevant at the extremes where it adds 4.75% to your chance of success, or avoiding failure.

Rolling dice is always fun so there is that, but it's not actually as useful as you might think.

Ed: the main benefit in practice, as I pointed out in my first post, is re-rolling failed saves.

[A_S]

My maths is correct — we did lots of calculations in a thread on this very site a couple of years ago; albeit we were looking at the feat which allows you to re-roll initiative and comparing that to Improved initiative.

I am aware that it squares the error probability but that said: At the level you get to cast 9th level spells it's only likely to be relevant at the extremes where it adds 4.75% to your chance of success, or avoiding failure.

Rolling dice is always fun so there is that, but it's not actually as useful as you might think.

Ed: the main benefit in practice, as I pointed out in my first post, is re-rolling failed saves.
First off: Your math actually isn't correct. The expected value of taking the better of 2d20 is 13.825, or an increase of 3.325 over 1d20's average roll of 10.5 (substantially higher than the +2.5 you mentioned earlier). You can find this number various places like here (http://rpg.stackexchange.com/questions/14690/how-does-rolling-two-dice-and-taking-the-highest-affect-the-average-outcome) with a bit of digging around, but if you want to prove it to yourself, here's a quick Python script that rolls 2d20 a million times, takes the higher each time, and averages the results:
from random import randint

results = []

for i in range(1000000):
a = float(randint(1,20))
b = float(randint(1,20))
results.append(max(a,b))

print sum(results)/len(results)
However, that's largely tangential to my point. As I said in my previous post, expressing the benefit of rolling two dice as an increase in mean result, even if done correctly, is misleading. If I want to know how much benefit I'm going to get out of rolling two dice on a D&D-style succeed-or-fail task, what I want to know isn't "how much higher am I going to roll on average," (which is useless information, because rolling higher on a d20 isn't better unless it makes the difference between success and failure). It's "how much more likely am I to succeed."

You say that this benefit will only be relevant at the extremes, but that's demonstrably not true. As I already mentioned, for a task where you have a 50% chance of success on one die (as far from the extremes as you can get), rolling two gives you a 75% chance of success. How is being half again as likely to succeed not relevant?

Perhaps what you mean is that at high levels, characters' bonuses in their areas of expertise are so high (and their numbers in other areas so comparatively low) that almost every check they make will either be nearly guaranteed (only fail on a 1) or nearly impossible (only succeed on a 20). However, even if that's what you mean, the benefit of rolling two dice is still relevant even aside from avoiding failed saves; the benefit on nearly impossible rolls, which you describe as "add 4.75% to your chance of success" is, in fact, nearly [i]doubling your chance of success. If you are really rolling checks where you need a natural 20 that often, then that's extremely relevant.

If what you're saying is that high level characters only ever roll checks where they only fail on a natural 1 (the only case I can think of where "preventing failed saves" is the only relevant part of getting to roll two dice), then may I suggest that your DM's may not have been appropriately challenging your party?

-----

There is a different argument against Choose Destiny as a spell that I find more convincing. It's something like, "if you are casting 9th level spells, you should be using them to win entire encounters without having to bother making any actual rolls to do so; with that in mind, even a substantial increase to your odds of success on your rolls is underwhelming for a spell of this level." That argument, to my mind, has some merit; the bar for the power of 9th level spells is awfully high.

But it seems blatantly fallacious to me to argue that the benefit of rolling two dice on all your d20 checks is just rerolling failed saves and having fun rolling dice.

[ben-zayb]

My typical unseen seer build looks like:
Human
Spellthief 1 / Wizard 5 (with spontaneous divination) / Incantatrix 3 / Unseen Seer 5 / Prestige Bard 1 / Unseen Seer +5You do realize you're only getting 0-level Bard Spells from that in exchange for 1 caster progression loss, right?

Another advice to the OP, especially if you have spontaneous divination, is to nab Alter Fortune (it's a Wizard spell, so it won't take up your free Unseen Seer spells) and never look back.

[thethird]

I get +2 to divination levels and all bard spells added to my spell list.

[ben-zayb]

I get +2 to divination levels and all bard spells added to my spell list.

That's contrary to the RAW on prestige classes, though. Bard 1 gets 0-level bard spells, which are then added to your spell list upon taking Prestige Bard 1.

[thethird]

I've always read, played and ruled this:

"The bard, paladin, and ranger spell lists contain a number of spells that don't appear on other classes' spell lists. In general, any character who enters one of these prestige classes should gain access to spells unique to that class's spell list, at the same levels indicated for the standard class."

With the last levels (bolded) being spell levels and not class levels. But upon rereading it I can see where you are coming from.

[Flickerdart]

That's contrary to the RAW on prestige classes, though. Bard 1 gets 0-level bard spells, which are then added to your spell list upon taking Prestige Bard 1.
This interpretation would mean that prestigious classes never earn their highest-level spells, which goes against the whole point of having them.

[ben-zayb]

This interpretation would mean that prestigious classes never earn their highest-level spells, which goes against the whole point of having them.

Aside from getting nearly every other class feature while progressing your own spellcasting partially. Seems like a fair trade by WoTC standards (not ours).
:smallbiggrin:

[Flickerdart]

Aside from getting nearly every other class feature while progressing your own spellcasting partially. Seems like a fair trade by WoTC standards (not ours).
:smallbiggrin:
Nearly? far as I can tell they get everything, and I don't see why people who take the classes should be cheated out of high-level spells due to a twisted reading.

[ben-zayb]

Nearly? far as I can tell they get everything, and I don't see why people wjho take the classes should be cheated out of high-level spells due to a twisted reading.

Cheated out of practically gestalting
EDIT: including "nearly" and "other" was a redundancy mistake.

[Flickerdart]

Cheated out of practically gestalting
Um...no. Not even close. But even if these classes were that powerful, that doesn't change the letter of the rule, nor its spirit.

[ben-zayb]

Um...no. Not even close. But even if these classes were that powerful, that doesn't change the letter of the rule, nor its spirit.

You don't change the letter of the rule when it's a RAW "interpretation", as you said earlier.
It's in the "ask your DM" category, along with the Tumble vs Thicket of Blades argument.

Now, if you could point me to any statement from the creators of this prestige class, we can discuss the "spirit" of the letter, or what "goes against the whole point of having them".

Anyway, this is really getting off topic. You really provided no RAW counterpoint as to why this isn't valid RAW interpretation, and objecting just based on what you think is or isn't RAI. I'd suggest we just agree to disagree.

[Coidzor]

That's contrary to the RAW on prestige classes, though. Bard 1 gets 0-level bard spells, which are then added to your spell list upon taking Prestige Bard 1.

Things are bad enough when it comes to interpretting Unearthed Arcana. We don't need people trying to foist spurious readings of RAW on others as if it were absolute fact. :smallconfused: